Evaluation of the user/operator fatigue using eye-tracking with machine learning algorithm

Abstract

The study of fatigue through eye tracking has been an intriguing topic done by researchers from many parts of the world. Many parameters were studied to get a better visualisation and detection of fatigue in the early stages. This is to anticipate and prevent future accidents from happening. These studies helped us get a better understanding on how fatigue occurs and ways to detect them through physical and psychological symptoms. This project aims to observe and study the pupil velocity and size, gaze position and accelerometer of the eyes through machine learning. This will be done by training data through classifiers and analysing the results to achieve detection of fatigue. The results will be compared with different classifiers available in Matlab and compared with the data that has been obtained through an experiment held prior to this project.